unused code: task error handlers

[whoabuddy]

across the job management / task system each class has _should_retry_on_error and _handle_execution_error with some code but are unused within the tasks themselves

the execution errors can be a sink for anything that happens at the Python level since we expect a success: False from that object when something goes wrong

and now that we have strict parsing of ts tool results we can use should retry conditions to check things like BadNonce, ConflictingNonceInMempool, NotEnoughFunds, or other common broadcast errors from Stacks.js